Handheld meter robustness testing

[GuidoK]

Some time ago a few others had asked about looking at Benning

I assume in that case you're thinking about the Benning M12, their top model?
https://www.benning.de/products-en/testing-measuring-and-safety-equipment/digital-multimeter/logging-multimeter-mm-12.html
4k/40k count, bluetooth logging, ac+dc, dual display, dBm
That is a rebrand of the Appatech Appa 506B:
https://www.appatech.com/en/product-553883/APPA-500-SERIES-MULTIMETERS-APPA-506-APPA-506B.html

Appatech also has the Appa 505, a 10k/100k count meter, also sold under the name ISO-TECH IDM505
I think this was a popular meter here on the forum due to a good pricepoint at some time.

and Testo.

Looking at their meters, I'm not really blown away by their specs. It looks like all their meters are more targetted towards electricians/industry than electronics.
Then again, if that is their target market, the electrical robustness must be good.

[GuidoK]

I should probably add the BM857S to the shortlist because that's the one I own. 

But that would be boring.
Before the tests start we already know it will be robust, there won't be any quirks, it'll be reliable and there will be no monumental screw ups in the design

[Fungus]

But that would be boring.
Before the tests start we already know it will be robust, there won't be any quirks, it'll be reliable and there will be no monumental screw ups in the design

OTOH: Millions of people will see that Brymen makes square, boxy, "industrial" meters, too.

eg. One of the reasons I chose the BM857 is because there's almost no function overloading on the dial. People go on and on about the Fluke87V's simple user interface but this is even better. In a way the 857/859 are Brymen's most direct competitors to the 87V.

[EEVblog]

eg. One of the reasons I chose the BM857 is because there's almost no function overloading on the dial. People go on and on about the Fluke87V's simple user interface but this is even better. In a way the 857/859 are Brymen's most direct competitors to the 87V.

The 87 has never had a "simple" user interface. It has three different power up modes like centre zero bargraph mode etc.

[Fungus]

eg. One of the reasons I chose the BM857 is because there's almost no function overloading on the dial. People go on and on about the Fluke87V's simple user interface but this is even better. In a way the 857/859 are Brymen's most direct competitors to the 87V.

The 87 has never had a "simple" user interface. It has three different power up modes like centre zero bargraph mode etc.

Doesn't stop people going on and on about it though.

(I'm guessing most of them don't know about that feature or if they do they never use it)

[joeqsmith]

Some time ago a few others had asked about looking at Benning

I assume in that case you're thinking about the Benning M12, their top model?
....

I didn't record the model numbers and have not looked at what they offer.  More just a reminder.   

I was not impressed with Keysight's lack of communication and how poorly the product performed but we could have a look at one of their high end meters.   

[GuidoK]

I was not impressed with Keysight's lack of communication and how poorly the product performed but we could have a look at one of their high end meters.

That's also very interesing. Maybe more people here are contemplating whether or not to buy a high end keysight meter
That cheap keysight indeed did not perform well on multiple aspects: electric robustness wasn't very good, the indentation spring from the knob completely failed (which looks like something they just didn't test; why use a different plastic when you can just use a plastictype you know that works as millions of meters use that), and the contact pads/springs from the knob were very worn after the lifecycle test (that surprised me very much...apparently not all pcb contactpads and springs are equal...)
Testing the a higher end meter would reveal whether Keysigth can't make a robust meter or that they just skimp on the quality of the cheaper models.
And that is an interesting comparison to Fluke (I think that is keysights main target market): the cheaper lower end Flukes are just as robust and have the same quality as the higher priced high end Fluke models (the china made 17b+ still has the best rotary knob in your test I believe?)

Although a 'cheap fluke' is still relative of course...where I live the Fluke 115 is still more expensive than a Brymen BM869S with silicone test leads and soft case...

[jspencerg]

I would like to see a high end meter tested if it were a new release with some innovations/ improvements for the portable meter market.  However,  watching you torture and rebuild the UT61E&E+ was most entertaining and educational.  I'd like you to run the repaired E+ to failure(12Kv?), get a replacement to test.  Test the virgin E+ and remove the asterisks you've placed on the E+ results.

[joeqsmith]

Added it.   Keep in mind that I have ran more UNI-T products than any other brand and they never perform well.

[GuidoK]

So I stumbled upon this video on youtube showing the Brymen BM869S to be somewhat more susceptible from interference when measuring high resistance compared to other meters. The same way (with the hand) as for instance the Gossen Metrawatt was susceptible in the mV range.
I also did this test and indeed my BM869S also has this characteristic in considerable larger amount than a fluke 115 and an Aneng an8009 I compared it with.

And analyzing your video from the BM789 I think the new BM7## series might also be susceptible to it.

It's not a big problem for me as I never really measure such high resistance (I didn't even have a resistor that high so had to put a few in series), but I am interested in where this quirk comes from and if it's possible to correct within reasonable means (maybe add some shielding over a component or so).
Is it maybe a bespoke ADC that brymen uses that is more susceptible?
I could not detect the BM869s for example using significantly less current when measuring this high resistance compared to the other meters.

I know Dave once mentioned it in a video that when measuring high resistance the electric field radiating of your body can be picked up by the wiring because the current flowing at such measurements is tiny, but the difference between the BM869S and a few other meters is there.

It might be an interesting idea for a short video?

This was the video I saw on youtube:
https://youtu.be/Ehy_uZygdFA
(I think it's heavily inspired on your Gossen Metrawatt video  )

[joeqsmith]

I was expecting the video showing the problem with 50Hz line.   The one you linked doesn't surprise me at all.   I make a habit of staying away from the setup when measuring low voltages, low currents, high resistance, etc to avoid such problems.    I frequently show the meters reading 40M and will talk about the stability. 

I like the Fluke 189.  I think it has 3 digits and will move 2 digits with a wave of my hand.

Recently there was a person posting about measuring a carbon resistor and how much it was drifting.   I made an attempt to replicate their tests and showed the effects of my walking into the lab, and then using a decent resistor.

[GuidoK]

Recently there was a person posting about measuring a carbon resistor and how much it was drifting.   I made an attempt to replicate their tests and showed the effects of my walking into the lab, and then using a decent resistor.

I understand that, but in this case it's not the resistor but the meter itself.
The difference in susceptibility between the BM869S and a lot of other meters is at least an order of a magnitude if I were to guess, maybe even 2 in extreme case depending on what control meter is used.
That is of course not as much as that quirk in the Gossen metrawatt (there it was at least 3 orders of mangitude in the extreme case I think), but still the difference is there.

[joeqsmith]

For fun, I repeated their test using the BM869s that I used for life cycle testing.  This meter has the later firmware which I expect would be closer to the OPs).     I also used one of my Fluke 189.   Both meters on the bench, sitting on a grounded ESD mat.     I tried with and without leads.  Leads were 27" long.   I left them untwisted laying across the mat.  I used that Caddock USVD2 2ppm 20.0007Meg resistor that is mounted in the foam as my reference.  This part has the banana connectors and will plug directly into meters that use the standard spacing.   Meter were set to record their min/max.  I then waved my hands around the meters and also got up from the chair a few times (maybe 2 feet away from the meters).  Keep in mind that we are in the middle of winter and the house is dry which it great for ESD.

 
With the leads:
Fluke 189  Max: 23.256M, Min: 7.772M
Brymen BM869s Max: 22.275M, Min: 17.800M

Without the leads (resistor plugged directly into the meters):
Fluke 189  Max: 20.032M, Min: 19.682M
Brymen BM869s Max: 20.134M, Min: 19.899M

Are both of these meters junk?  Maybe but I rather like them.   I suspect that the Fluke they show has a much higher voltage with their resistor attached than their Brymen.  For the people wanting to look at LEDs with their $400 meters, the 87V may be a better choice. 

***
Added picture of the test resistor used.  Also, I want to be VERY clear.  The test I ran is not controlled and I am NOT suggesting that the Fluke 189 is more susceptible than my Brymen BM869s! 

***
Because I opened the question of the test voltage, using the same 20M resistor, I measured the voltage across it with my old electrometer.  Obviously, at 20M we need the high input impedance.   I tried my 87V that I purchased brand new a couple years ago specifically for testing.  It's seen some abuse but I doubt it would have any effect on this measurement.

Fluke 87V:  3.590V
Brymen BM869s: 0.735V
UNI-T UT61E+: 0.814V
UNI-T UT181A: 0.809V

Of the four meters, it would make sense that the Fluke 87V would have better immunity in this particular case but  again, this is not how I would use the meter and I'm not a fan of these high test voltages (for electronics work).  But, if I needed to light some LEDs, this would be the ticket.

[GuidoK]

  I suspect that the Fluke they show has a much higher voltage with their resistor attached than their Brymen. 

Maybe, I don't have an 87V, but I doubt that that is the cause/explanation of this effect, at least not what I'm seeing with my meters. (that klein meter in the video is also not affected)

Like I said, I've measured the current. Of course this current is very very low (under 0.10µA), so the numbers I measure are not very accurate in an absolute way, but it does show their relative value to eachother (and had both the BM869s and an8009 in series to see if there are large differences in the µA range).
My fluke (115) uses about as much current as the BM869s and is noticeably less susceptible (also with min/max), and my an8009 is probably least susceptible (it has no min/max function, but looking at the display the difference is quite big so I'm pretty confident in that), and with my current measurement it uses slightly less current than the BM869s (the difference is not big, but it's there; maybe it has a slightly different internal resistance).
They're imho certainly not large differences that could explain the difference in susceptibility due to using considerably higher voltage for the resistance measurement.
But maybe I'm measuring incorrectly. I don't have an electrometer. I would think that the current measurement is also ok and that Ohms law also works with these high resistances.

I don't get the big min/max differences you get (I can only measure with leads) but where I live it's probably less cold and less dry.

[joeqsmith]

Check my last updated post for the voltage measurements.   The 87V would have the highest current at 3.59V/20M = 179.5nA.     

No doubt that there are other factors that would play into it.  I would expect meters with heavy filters would be less sensitive.   I don't record that data but we certainly have looked at it.  It seems I had ran a test once with the 121GW where I was modulating the resistance mode to get a feel of how their filters behave.  Heavy filter may mean slower response.  Fast response, like the original UT61E may mean more sensitive to ESD.   In the end, it's all a tradeoff.  That Gossen for example has the highest performance of any handheld meter I have looked at in some areas.   

I wouldn't be surprised if the companies didn't pick different filters (for example) based on the expected use.  The 87V may have been targeted for the industrial technicians where the 189 may have been more for the engineering group.  I certainly have no problem with the 189.     

It's up to the buyer to pick the tools for their particular use.  There's lots of options and the free meters from HF have the longest battery life I have measured and, they are FREE.  lol.  It's also about as good as my first Fluke.   

I am not surprised you wouldn't see the same results with your min/max test.  Again, it's not a controlled test.   

If your goal is to improve the BM869s, it would be best to open a discussion with Brymen.

[GuidoK]

It's up to the buyer to pick the tools for their particular use. 

.....


If your goal is to improve the BM869s, it would be best to open a discussion with Brymen.

No my goal is not to improve the bm869s. Like I said I never work with these high values. I work more in the mV ranges and bought the meter specifically for the 500k mode in the DC mV range combined with the high accuracy and the possibility to have it calibrated according to iso17025.
My post was more to bring it under the attention that the bm869s is more susceptible to this that a lot of other meters.

It's indeed up to the buyer to pick the tools for their use. But this kind of behaviour of course isn't advertized or specified, so if that's important to a buyer the only way to find out about it before purchase is if people write about it/document it. So maybe that was my goal

[joeqsmith]

It's up to the buyer to pick the tools for their particular use. 

.....

If your goal is to improve the BM869s, it would be best to open a discussion with Brymen.

No my goal is not to improve the bm869s. ...   So maybe that was my goal

Your first comment. 

...but I am interested in where this quirk comes from and if it's possible to correct within reasonable means

[GuidoK]

I ment that in a hypothetical way.
The process and means of whats involved to potentially correct it and what causes it interests me.

I never really measure high resistances so I have no need for it. (And I'm not going to modify anything inside my meter from a safety standpoint)

But I might contact Brymen
See if they are aware of this behaviour and know how it causes the meter to behave like this

Edit:I might have misread your post about improving the bm869s; I thought you ment the bm869s I have here on my bench, so my bm869s (I'm not gonna modify that one), but you probably ment the bm869s in general, so the design of it. Then yes, why not

[joeqsmith]

Yes, I meant in general. 

[Kosmic]

I used that Caddock USVD2 2ppm 20.0007Meg resistor that is mounted in the foam as my reference.

Small detail, but 2ppm is the TC of the ratio between the 2 resistors in the divider. Absolute TC is 10ppm/C.

[cybercorfu]

Found the problem it was a ceramic 20uF capacitor not completely broken just with 40kΩ  resistance causing semi short to ground

[GuidoK]

To celebrate reaching 20k subscribers, the following list of handheld meters will be considered. 

I think this meter might also be interesting to test:
AmazonCommercial 90DM610
https://www.amazon.com/AmazonCommercial-Count-Digital-Multimeter-CATIV/dp/B07W1BL3RH

This falls I think in the popular Brymen BM235 / Fluke 117/keysight u1230 class, so the "quality entry level" class I guess.
It's a rebranded CEM-9562
https://www.cem-instruments.com/en/Product/detail/id/1434
Independently tested and it has quite an elaborate input protection network (5 MOV's, 4 PTC's I believe). IP67 too (I don't know if that's a plus though)
But it's crazy cheap, $40 (in europe it's more expensive, €60), so even well under the price of the BM235 (I think about half that?).
I wonder if its as robust as it looks. I think it would also be interesting to put the rotary switch through the lifecycle testing. See if it's either fluke like quality or if it's keysight quality 

[joeqsmith]

That last video kicked the channel over the 20k mark..  Crazy!!   So as promised,  I plan to start the poll in a week or so.  In the mean time if you want to see some other meter ran that is not on the current list, feel free to chime in.

https://www.eevblog.com/forum/testgear/hear-kitty-kitty-kitty-nope-not-that-kind-of-cat/msg3976493/#msg3976493

[theHWcave]

Hi Joe, congrats in reaching 20K subs. Awesome and well deserved.

Please add the low budget KAIWEETS KM601 Smart meter. They sent me one and I did a review (it has a few problems). Inside, I thought its fuses are better than most budget meters but the rest of the protection is hardly worth mentioning. Then again the similar underwhelming AN8008 was doing quit well if I recall in your tests.  Anyway, checking with KAIWEETS regarding their CAT ratings, they stated their meters are tested by an independent Chinese testing lab, called NTEK. That lab's website shows an impressive list of international certificates. I know you are not checking CAT testing but I would be really interested in how well it stands up to your transient tests.
The KM601 is a smart meter and from the comments I got I think it is quit popular with beginners because it looks like a phone = familiar instead of of "complicated" like a multimeter, and of course it automatically switches between measuring volts, resistance and continuity, arguably the most common uses.

[joeqsmith]

Poll is now open. 

Please add the low budget KAIWEETS KM601 Smart meter.

Added.